[Letux-kernel] [PATCH RFC] net: hso: register netdev later to avoid a race condition

Andreas Kemnade andreas at kemnade.info
Wed Apr 26 19:22:58 CEST 2017


Hi,

On Wed, 26 Apr 2017 18:15:10 +0200
"H. Nikolaus Schaller" <hns at goldelico.com> wrote:

> Hi,
> 
> > Am 26.04.2017 um 18:03 schrieb Andreas Kemnade <andreas at kemnade.info>:
> > 
> > On Wed, 26 Apr 2017 08:53:18 +0200
> > "H. Nikolaus Schaller" <hns at goldelico.com> wrote:
> > 
> > 
> >>> But nevertheless I managed to reproduce the kernel oops.
> >> 
> >> Ok, that is the most important thing for kernel debugging.
> >> 
> > Hmm, I checked again.
> > - reproducible with 4.10 without your hack and without "net: hso:
> > register netdev later to avoid a race condition"
> > - reproducible with 4.11rc8 without you hack and without "net: hso:
> > register netdev later to avoid a race condition"
> > - not reproducible with 4.11rc8 without your hack but with "net: hso:
> > register netdev later to avoid a race condition"
> 
> > 
> > hmm, am I not patient enough to have it crash? Or have you mixed up
> > this patch with my earlier patch about the rmmod problem?
> 
> Ah! You are right. Looks as if I mixed it up with the rmmod problem which
> I had not seen before... I think I am getting old...
> 
The rmmod problem fix seems to go into 4.11 if I understand the
branching system correctly. 

> So I never tried the patch that you did include in your first mail on this
> thread. Looks as if there were too many other patches queued up before I found
> time to look into them and then e-mail isn't the best "todo" archive.
> 
> Maybe a better strategy would be to pile them all up in some work/kemnade/hso
> feature branch. Then none gets lost. This is a little different from LKML
> patches where there is patchwork.kernel.org to collect all patches and track them.
> 
If it is not going upstream in a fast way, I will add such a branch.
This time I send a patch rfc explicitly because I could not reproduce
the problem.

> > Just asking to avoid debugging a possibly non-existent problem.
> 
> I will try it asap and report if it is equally good as my mdelay hack...
> 
I will send the patch upstream. You can then add a tested-by. If
that patch does not fix things, I am out of ideas anyways, so we have
to ask upstream anyways. Then there are chances that it will be accepted
before the next merge window opens.

Regards,
Andreas
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 819 bytes
Desc: OpenPGP digital signature
URL: <http://lists.goldelico.com/pipermail/letux-kernel/attachments/20170426/a0601406/attachment.asc>


More information about the Letux-kernel mailing list